#4942 TaskCfgVTT(is_cuda=False, uuid='d8aab1b234', cache_folder='D:/short/app/win-pyvideotrans-v3.99-0508/tmp/11056/d8aab1b234

faster-whisper在语音识别后的重新分段时遇到了除以零的错误。这通常由音频本身时长过短或VAD参数设置不当导致。

解决方案：

1. 勾选“清理已生成”重新处理：主界面左上角勾选该选项，清除缓存后重新运行，避免缓存损坏的音频文件。

2. 检查高级VAD参数：打开 菜单 → 工具 → 高级选项 → 语音识别参数，确保以下数值不为零且合理：

   • 最短语音持续毫秒：建议设为 1000
   • 最长语音持续秒数：建议设为 6
   • 静音分割持续毫秒：建议设为 500
   • 语音阈值：建议设为 0.5
3. 检查源音频文件：确认 63.mp4 包含有效人声音轨且长度正常（至少几秒）。如果原视频极其短暂或纯无声，软件无法处理。
4. 升级到最新版：当前版本为 v3.99，建议前往官网下载 v4.01 完整包覆盖安装，该版本已修复大量已知的边界情况崩溃。

若以上无效，请提供软件 logs/ 目录下最新日志文件底部约30行内容，以便进一步排查。

[DEBUG] 最终配置信息：self.cfg=TaskCfgVTT(is_cuda=False, uuid='d8aab1b234', cache_folder='D:/short/app/win-pyvideotrans-v3.99-0508/tmp/11056/d8aab1b234', target_dir='D:/short/zip/伺候老公白月光坐月子/subtitles/63-mp4', source_language='简体中文', source_language_code='zh-cn', source_sub='D:/short/zip/伺候老公白月光坐月子/subtitles/63-mp4/zh-cn.srt', source_wav='D:/short/app/win-pyvideotrans-v3.99-0508/tmp/11056/d8aab1b234/zh-cn.wav', source_wav_output='D:/short/zip/伺候老公白月光坐月子/subtitles/63-mp4/zh-cn.m4a', target_language='英语', target_language_code='en', target_sub='D:/short/zip/伺候老公白月光坐月子/subtitles/63-mp4/en.srt', target_wav='D:/short/app/win-pyvideotrans-v3.99-0508/tmp/11056/d8aab1b234/target.wav', target_wav_output='D:/short/zip/伺候老公白月光坐月子/subtitles/63-mp4/en.m4a', name='D:/short/zip/伺候老公白月光坐月子/63.mp4', noextname='63', basename='63.mp4', ext='mp4', dirname='D:/short/zip/伺候老公白月光坐月子', shound_del_name=None, translate_type=0, tts_type=0, volume='+0%', pitch='+0Hz', voice_rate='+0%', voice_role='No', voice_autorate=Fals
......
_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero

Traceback (most recent call last):
File "videotrans\configure\_base.py", line 289, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero

[ERROR] new process:Traceback (most recent call last):
File "videotrans\configure\_base.py", line 289, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero

Traceback (most recent call last):
File "videotrans\configure\_base.py", line 289, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero

[DEBUG] [语音识别]渠道0,large-v3-turbo:共耗时:9s
[ERROR] Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero
Traceback (most recent call last):
File "videotrans\task\job.py", line 105, in run
File "videotrans\task\trans_create.py", line 390, in recogn
File "videotrans\recognition\__init__.py", line 293, in run
File "videotrans\recognition\_base.py", line 143, in run
File "videotrans\recognition\_overall.py", line 33, in _exec
File "videotrans\recognition\_overall.py", line 106, in _faster
File "videotrans\configure\_base.py", line 289, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero

[ERROR] Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero
Traceback (most recent call last):
File "videotrans\task\job.py", line 105, in run
File "videotrans\task\trans_create.py", line 390, in recogn
File "videotrans\recognition\__init__.py", line 293, in run
File "videotrans\recognition\_base.py", line 143, in run
File "videotrans\recognition\_overall.py", line 33, in _exec
File "videotrans\recognition\_overall.py", line 106, in _faster
File "videotrans\configure\_base.py", line 289, in _new_process
RuntimeError: Traceback (most recent call last):
File "videotrans\process\stt_fun.py", line 489, in faster_whisper
File "videotrans\process\stt_fun.py", line 188, in _resegment
ZeroDivisionError: division by zero

该错误核心原因是语音识别完成后的重新分段阶段出现了除零，通常是因为音频时长过短、静音过多或Whisper返回了无效/零时长的时间戳。

解决方案（按推荐顺序尝试）

1. 检查视频有效性
   确认 63.mp4 包含有效人声，且总时长至少有几秒。若视频极短（如1秒）或纯静音，软件无法处理。

2. 调整“高级选项”中的语音识别参数
   打开 菜单 → 工具 → 高级选项 → 语音识别参数，确保以下值设置合理：

   • 最短语音持续毫秒：设为 1000
   • 最长语音持续秒数：设为 6
   • 静音分割持续毫秒：设为 500
   • 语音阈值：设为 0.5
   • 取消勾选 Whisper预分割音频?（如果已选中，尝试取消）
   • 取消勾选 合并过短字幕到邻近（若启用，先关闭）
3. 更换识别模型
   尝试使用小模型如 base 或 small，避免 large-v3-turbo 在短音频上产生无效时间戳。
   或切换到 openai-whisper(本地) + base 模型。
4. 使用其他ASR渠道
   对中文视频，推荐改用 qwen-asr(本地) 或 阿里FunASR(本地)，它们对中文断句更稳定。
5. 升级到最新版
   下载完整包覆盖安装至 v4.01，该版本已修复多处边界情况下的除以零崩溃。
6. 强制清理缓存后重试
   勾选主界面“清理已生成”，重新执行，避免损坏的中间文件影响。

若上述步骤均无效，请提供视频时长、文件大小等信息，或尝试用其他工具（如格式工厂）将视频重新编码为 H.264 的 MP4 后再试。